1. Field of the Invention
The present invention relates to information processing systems, information processing apparatuses, and computer-readable recording media used in the systems or the apparatuses, suited to cases in which electronic-watermark information is embedded in input information, such as digital image data or digital sound data, to protect copyright, to prevent forgery, and to record various types of information.
2. Description of the Related Art
As computers and networks have been remarkably developed in recent years, various types of information, such as character data, image data, and sound data, has been handled in the computers and networks. Since such data is digital, it can be easily copied with its quality maintained. Therefore, to protect the copyright of such data, copyright information and user information are embedded in image data and sound data as electronic-watermark information (hereinafter just called an electronic watermark) in many cases.
With the use of an electronic-watermark technology, information which people cannot recognize with their senses of sight and hearing is confidentially embedded in image data or sound data. When an embedded electronic watermark is extracted from image data or sound data, copyright information, user information, and identification information are obtained, and illegitimate copying can be traced.
A first condition required for such an electronic watermark is a quality at which embedded information cannot be identified, namely, at which information is embedded in the original digital information such that the quality of the original digital information does not deteriorate.
A second condition is robustness with which information embedded in the original digital information remains, namely, with which embedded information is not lost even if editing, such as data compression and filtering, or an attack is applied.
A third condition is the amount of information to be embedded, which can be selected according to use.
These conditions, required for electronic watermarks, are generally tradeoffs to each other. When an electronic watermark having a high robustness is generated, for example, relatively large quality deterioration occurs and the amount of information to be embedded becomes small in many cases.
Methods for embedding electronic watermarks in multi-valued still pictures can be divided into two types, spatial-domain embedding methods and frequency-domain embedding methods.
In each of various electronic-watermark embedding methods, embedding processing corresponds to extracting processing one by one, and there are no compatibility. In general, it is said that spatial-domain embedding methods provide a low quality deterioration with a low robustness, whereas frequency-domain embedding methods provide a high robustness with a relatively high quality deterioration. Each method has a different feature, such as a high robustness with a small amount of embedded information, or a high quality with a low robustness.
To protect embedded information, information (hereinafter called a key) indicating algorithm, embedding positions, and changes are kept confidential in many cases. This is to enhance robustness to an intentional attack in which the algorithm and embedding positions are analyzed to remove electronic watermarks.
It can be considered to efficiently protect copyright that a monitoring organization for checking whether illegitimate copying is performed, by extracting electronic watermarks is provided. It is important for such a monitoring organization to keep the electronic-watermark method used and the key confidential to avoid an intentional attack.
As described above, there are various electronic-watermark methods each having its features. There are also many companies and associations which want to prevent illegitimate copying and illegitimate outputting of digital data by the use of electronic watermarks. If such companies and associations independently select electronic-watermark methods to embed electronic watermarks in data, since embedding processing and extracting processing correspond one by one in each of electronic-watermark methods and they are not compatible, the following problems occur.
1. It is difficult for one monitoring organization to perform united checking because electronic-watermark extracting processing is needed for each method.
2. To perform united checking by one monitoring organization, the monitoring organization needs to have all electronic-watermark extracting techniques, causing a large load.
3. The monitoring organization needs to manage the keys corresponding to all the extracting techniques confidentially and strictly.
4. When a monitoring organization is provided for each method, if an embedded electronic watermark cannot be extracted, it cannot be determined whether the embedded electronic watermark has been generated by a different method, or the electronic watermark is broken by an attack.
One monitoring organization refers to one configuration in terms of a system or a method, determined by standardization or a nation, rather than one physical organization.